[1]单强‘,孙晓明.多特征分层融合医疗设备图像检索方法[J].哈尔滨理工大学学报,2017,(02):135-139.[doi:10.15938/j.jhust.2017.02.025]
 SHAHanR-,SlXiao-minR-Z.Medicallmage Retrieval Method Basedon Multi Feature Hierarchical Fusion[J].哈尔滨理工大学学报,2017,(02):135-139.[doi:10.15938/j.jhust.2017.02.025]
点击复制

多特征分层融合医疗设备图像检索方法()
分享到:

《哈尔滨理工大学学报》[ISSN:1007-2683/CN:23-1404/N]

卷:
期数:
2017年02期
页码:
135-139
栏目:
计算机与控制工程
出版日期:
2017-04-25

文章信息/Info

Title:
Medicallmage Retrieval Method Based on Multi Feature Hierarchical Fusion
文章编号:
1007-2683(2017)02-0135-05
作者:
单强‘孙晓明2
(1首都医科大学附属北京朝阳医院,北京100020; 2哈尔滨理工大学测控技术与仪器黑龙江省高校重点实验室,黑龙江哈尔滨150080)
Author(s):
SHAH伪anR-Sl、 Xiao-minR-Z
(1. Beijing Chaoyan} llospital Affiliated to Capital Medical I miversitv ,Beijing 100020,China; 2. lli}her L,ducational Ke、 Laboratory for Measuring&Control ’fechnolo}y and lnstrumentation of lleilon}jian} Province llarbin Gniversity of Science and ’fechnolo}y, llarbin 150080,China)
关键词:
关键词:图像检索多特征分层融合医疗设备
Keywords:
Keywords:image retrievalmulti featurehierarchical fusionmedical equipment
DOI:
10.15938/j.jhust.2017.02.025
文献标志码:
A
摘要:
摘要:针对解决医疗设备图像的检索问题,提出了一种基于多特征分层融合的图像检索方 法。此方法采用颜色矩计算图像的颜色特征,采用LBP算子计算图像的纹理特征,采用Fourie:描 述子计算图像的形状特征,通过分层加权融合的方法将三类特征融合在一个相似性测度之内,最终 形成检索判断的依据。根据窥镜图像和CT图像展开实验研究,其结果证实了所提出方法的有效 性。随着查全率要求的不断提升,所提出方法的查准率明显高于三种单一特征的检索方法,适合于 医疗图像的检索。
Abstract:
Abstract:Objective:In order to solve the problem of medical equipment image retrieval,an image retrieval method based on multi feature layer fusion is proposed. Process:This method using color moments to calculate the characteristics of color image,the LBP Operator is used to calculate the texture features of the image,using Fourier descriptors calculated image of the characters,the stratified weighted fusion method of the characteristics of three types of fusion within a similarity measure,and ultimately the formation of retrieval based on. Results and conclusions:According to the experimental study of the endoscopic image and CT image,the experimental results confirm the validity of the proposed method. With continuous improvement of the precision requirements of the proposed method in precision ratio was significantly higher than that of two single feature retrieval method for medical image retrieval.

备注/Memo

备注/Memo:
收稿日期:2016 - 07 - 06 基金项目:国家自然利一学基金(61401126) ;黑龙江省自然利一学基金(QC2015083). 作者简介:单强(1982-) ,男,工程师. 通信作者:孙晓明(1982-),女,副教授,硕士研究生导师,L,-mail ; xiaomin}一6881982@ 163
更新日期/Last Update: 2017-06-13